You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Currently the JSON API only supports a single configurable value for CQL query timeouts, which defaults to 10s in the standard config file that is checked in.
At high scale usage in cloud environments, we've observed the need for more flexible controls for different types of queries. For example, DDL queries like create/delete table have different behavior and can take longer in cloud environments.
JSON API can make use of the execution profile feature offered by the CQL in order to create profiles for each type of query, with separate timeouts for each profile. This will allow users to have more flexible control over timeouts with a simple config file change.
Suggested profiles:
default (use for DML that doesn't include LWT)
lwt (use for DML that includes LWT)
schema (DDL such as create/delete table)
count (use for SELECT COUNT queries)
The text was updated successfully, but these errors were encountered:
We might also be able to use a count profile to disable CQL warning logs for COUNT statements only, to get rid of these annoying log messages
Query '[0 values] SELECT COUNT(1) AS count FROM my_namespace.my_collection' generated server side warning(s): Aggregation query used without partition key
Currently the JSON API only supports a single configurable value for CQL query timeouts, which defaults to 10s in the standard config file that is checked in.
At high scale usage in cloud environments, we've observed the need for more flexible controls for different types of queries. For example, DDL queries like create/delete table have different behavior and can take longer in cloud environments.
JSON API can make use of the execution profile feature offered by the CQL in order to create profiles for each type of query, with separate timeouts for each profile. This will allow users to have more flexible control over timeouts with a simple config file change.
Suggested profiles:
The text was updated successfully, but these errors were encountered: